glass in cultural heritage – an introduction

15
GLASS IN CULTURAL HERITAGE – AN INTRODUCTION António Pires de Matos 1,2,3,* , Augusta Lima 1,2 , Filipa Lopes 2 , Márcia Vilarigues 1,2 , Joaquim Marçalo 2,3 1. Departamento de Conservação e Restauro, FCT/Universidade Nova de Lisboa, 2829-516 Caparica, Portugal 2. Unidade de Investigação VICARTE – Vidro e Cerâmica para as Artes, FCT/Universidade Nova de Lisboa, 2829- 516 Caparica, Portugal 3. Instituto Tecnológico e Nuclear, Estrada Nacional 10, 2686-953 Sacavém, Portugal *[email protected] Abstract A brief description of different types of silicate glasses is made and the origin of colour in glasses is discussed. The principles of corrosion in glass, illustrated with a few examples, will be explained. Applications of several techniques in the chemical characterization of glasses will be shown, namely X-ray Fluorescence Spectrometry (XRF), Scanning Electron Microscopy (SEM), Optical Absorption Spectroscopy and Laser Ablation Fourier Transform Ion Cyclotron Resonance Mass Spectrometry (LA-FTICR/MS), and Particle Induced X-ray Emission (PIXE). Keyword: glass; cultural heritage; glass analysis; glass corrosion; colour in glass Introduction “A glass can be defined as “an amorphous solid completely lacking in long range, periodic atomic structure and exhibiting a region of glass transformation behaviour “ James Shelby [1] Figures 1 and 2 shows the non repetitive order at long range in silica glass and the transition temperature, respectively.

Upload: others

Post on 29-Mar-2022

0 views

Category:

Documents


0 download

TRANSCRIPT

Microsoft Word - Glass_Proceedings_Final version_1_3_2009.docAntónio Pires de Matos 1,2,3,*
, Augusta Lima 1,2
1,2 , Joaquim
Marçalo 2,3
1. Departamento de Conservação e Restauro, FCT/Universidade Nova de Lisboa, 2829-516 Caparica, Portugal
2. Unidade de Investigação VICARTE – Vidro e Cerâmica para as Artes, FCT/Universidade Nova de Lisboa, 2829-
516 Caparica, Portugal
3. Instituto Tecnológico e Nuclear, Estrada Nacional 10, 2686-953 Sacavém, Portugal
*[email protected]
Abstract
A brief description of different types of silicate glasses is made and the origin of colour in glasses is
discussed. The principles of corrosion in glass, illustrated with a few examples, will be explained.
Applications of several techniques in the chemical characterization of glasses will be shown,
namely X-ray Fluorescence Spectrometry (XRF), Scanning Electron Microscopy (SEM), Optical
Absorption Spectroscopy and Laser Ablation Fourier Transform Ion Cyclotron Resonance Mass
Spectrometry (LA-FTICR/MS), and Particle Induced X-ray Emission (PIXE).
Keyword: glass; cultural heritage; glass analysis; glass corrosion; colour in glass
Introduction
“A glass can be defined as “an amorphous solid completely lacking in long range, periodic atomic
structure and exhibiting a region of glass transformation behaviour “
James Shelby [1]
Figures 1 and 2 shows the non repetitive order at long range in silica glass and the transition
temperature, respectively.
Fig. 1. 2D representation of a silica glass.
Fig. 2. Variation of the specific volume with temperature during the solidification of crystalline
and vitreous material. Tg is the glass transition temperature of the amorphous material. Tm is the
melting temperature of the crystalline material.
The glass transition is transition in which a supercooled melt yields, on cooling, a glassy structure. Below
the glass-transition temperature the physical properties vary in a manner similar to those of the crystalline
phase. The transition is not as dramatic as the phase change that takes you from liquid to crystalline solids
Batch materials to make glass can be grouped in network formers (SiO2, B2O3, P2O5, GeO2, As2O3,
Sb2O3), fluxes/network modifiers (Na2CO3, Na2SO4, K2CO3, Li2CO3, PbO), network stabilizers
(CaCO3, Al2O3 , PbO), colourants and fining agents. PbO can be used as a flux or, if substituting the
silicon in the network, as network former. In Table 1 examples of compositions of different types of
glass are shown.
Bottle glass 1 Goblet
Na2O 17.5 15.22 2.6 3 - 10
K2O 5.48 8.9 0.4 - 1
CaO 29.22 3.8 13.31 0 - 1
MgO
PbO 32.5 0 - 10
3
Glazes used for ceramics can be considered as thin layers of glass. They are made by firing a
mixture of batch materials, being the composition defined to obtain a desired melting temperature.
In Table 2 are indicated the oxides which can be part of a glaze.
Table 2. Oxides used in glazes.
RO, R2O R2O3 RO2
Colour in glass
The colour in glass can be due to d-d electronic transitions, charge transfer between cations and
anions (e.g. Fe 3+
O 2-
), interaction of the light with small particles (Cu, Ag , Au and cadmium
sulfoselenide), interaction of the light with small crystals formed during the cooling of glass due to
devitrification (light diffusion – opaline glass) and optical interference in thin layers.
The d transition elements have d orbitals incompletely filled and the d electronic levels are identical
in energy for free ions. In a glass the metal ions are surrounded mainly by oxygen ions and the
interaction of the electric field causes a small splitting of the energy levels. The value of this
splitting depends on the field strength caused by the surrounding ions, the coordination number and
the geometric arrangement around the central ions. Fig. 3 exemplifies the ligand field splitting in
two complexes with different symmetry. The tetrahedral and octahedral coordination around Co 2+
ions in glass gives absorption bands around 530 nm and 680 nm respectively, being the glasses
formed rose and blue, respectively. An analogy with the complexes [Co(H2O)6] 2+
and [CoCl4] 2-
in
aqueous media with similar colours can be made. An interesting example of the influence of the
ligand field strength in the colour is given by Shelby [1] for a borosilicate glass with Co 2+
which is
dark blue purple. Substituting partially the oxygen anion by Cl - , Br
- or I
4
Fig.3. Ligand field splitting in tetrahedral and octahedral complexes.
When preparing a coloured glass using a d transition element it is important to define an adequate
composition which can affect the ligand field. The oxidation state of the cation can be controlled
with a reducing or oxidizing atmosphere in the furnace.
The d transition elements impart to the glass a varied palette of colours. A few examples are given
in Table 3.
Ion Coordination Colour
Cr 3+
Octahedral green
Cr 6+
Tetrahedral yellow
Fe 2+
Tetrahedral blue
Co 2+
Tetrahedral blue
Co 2+
Octahedral rose-purple
Cu 2+
Octahedral blue
When characterizing a glass by optical absorption spectroscopy it is important to obtain the
spectrum values for the UV, Visible and Near Infrared range. Fig. 4 shows the absorption zones and
the respective responsible phenomenon. Intense bands with a maximum in the UV range can give
origin to colour as the band spreads to the visible range.
dxy, dxz, dyz
degenerate d orbitals
dxy, dxz, dyz
5
6
5
4
3
2
1
0 100 200 300 400 500 600 700 800 900 1000 1100
log
[nm]
a
b
Fig. 4. a) Absorption bands of the ligand or due to electronic transitions between the ligands and
the central cation; b) absorption bands due to transitions of d electrons between different energy
levels.
The colour can also be due to interaction of the electromagnetic radiation with small particles: Ag
(silver yellow), Cu or Au (ruby copper or ruby gold), cadmium sulfoselenides (red glass).
When CdSexTe1–x particles are introduced into glass batches, the pigment currently used for making
red glasses, the glasses formed contain semiconductor particles. The absorption spectra depend on
the concentration and size of the nanoparticles. Photons with energies greater than the band gap of
the semiconductor undergo absorption and consequently the optical absorption spectra have a sharp
transmission cut-off. The bands gaps of CdSe and CdS are 1.67 eV (724 nm) and 2.53 eV (490 nm)
respectively. The colour can be controlled by changing the ratio of the two compounds, varying
from yellow to dark red. A spectrum of a glass goblet containing sulfoselenide is shown in Fig. 5.
Fig. 5. Optical absorption spectrum of the red stem of a goblet from Fábrica-Escola Irmãos Stephen,
Glass Museum of Marinha Grande.
nm
abs
0
0,5
1
1,5
2
2,5
3
6
In the gold and copper ruby glasses as in the yellow silver staining in stained glass, the colour is due
to the interaction of the electromagnetic radiation in the visible range with metal nanoparticles,
originating absorption bands around 530 nm for gold, 565 nm for copper and 410 nm for silver.
These bands can be considered as plasma resonance bands as the free electrons in the particles
behave as a bounded plasma [1,4].
One of the first written documents where there is a recipe for making ruby glass was published by
Antonio Neri in his famous book ‘‘L’Arte Vetraria’’ [5]. Later, a report written by Kunckel
describes the manufacture of gold ruby glass [6]. Since then, gold ruby glass was produced in
several places in Europe. This production seems to have started in the 17th century when ‘‘Purple
of Cassius’’ was added to the glass batch compositions [7]. Purple of Cassius is a colloidal
suspension of gold obtained by reducing a gold (III) chloride aqueous solution with stannous
chloride. Fig. 6 shows a goblet made of ruby gold glass from the Glass Museum of Marinha
Grande.
Fig. 6. Bottle of ruby gold glass, (prob.) Marinha Grande, 19th century and its optical absorption
spectrum showing an absorption band with a maximum at about 530 nm.
In this case the gold could not be detected by XRF as the flask was made of lead crystal. Being lead
a heavy element, it absorbed the characteristic X-ray emission of gold as its concentration in glass
was certainly very low, probably around 0.01wt%. The absorption spectra unequivocally indicated
the presence of gold.
Chemical corrosion of glass
The chemical corrosion of glass is due to the contact of water (vapour or liquid) with the glass
surface causing alkali/alkaline earth species exchange with H + (H3O
+ ) [8] .
nm
abs.
7
Another type of exchange is shown in the following reaction
ªSiO - + Na
-
In any case, a layer of the surface of the glass is depleted of alkali.
The presence of OH - ions hydrolyses the ªSi-O-SiOª bonds.
ªSi-O-Siª + OH - ªSiOH + ªSiO
and a silica gel layer is formed.
The contact of historical glasses with a humid environment causes the alteration phenomenon called
crizzling which consists in fine cracks developed in glass as a consequence of poorly calculated
batch, usually with an excess of alkali and insufficient stabilizer [9,10]. Archaeological glasses
suffer chemical attack by natural agents and the weathering can be very destructive.
An example of the attack of glass objects by water is shown in Fig. 7.
Analytical techniques used for the characterization of glass - examples
Non destructive techniques that can be used for glass characterization are listed below:
Micro-EDXRF (energy dispersive micro X-ray fluorescence spectrometry)
UV-Vis-NIR (ultraviolet, visible and near infrared spectroscopy)
LIBS (laser induced breakdown spectroscopy)
ESEM (environmental scanning electron microscopy)
PIXE (proton induced X-ray emission) - with external ion beam
LA/ICP-MS (laser ablation inductively coupled plasma mass spectrometry)
Radiation measurement (for uranium and potassium analysis)
It is important to refer that the use of a laser leaves a very small mark not visible at naked eye.
As far as we know, there are two microdestructive sampling techniques used for glass analysis. One
uses a diamond tip to take a very small amount of glass (a few micrograms) [11]. If the glass has
some stress, there is the danger of breaking. The second method developed by Lopes et al [12] uses
8
a drop of diluted HF to dissolve the glass. The acid makes a very small mark so it should be applied
only in a part of the artwork not visible when exhibited. In both methods the chemical analysis can
be made either by ICP-MS or neutron activation analysis.
Fig.7. Attack of glass by water a) Crizzling, beginning - cloudy, hazy surface of glass; b)
Crizzling, beginning - cloudy, wet, hydrated surface of glass (“sweating" or "weeping"); c)
Crizzling, "full-blown", uniform cracking of the surface; d) Crizzling, advanced, surface beginning
to flake, or exfoliate.
(courtesy of Stephen Koob of The Corning Museum of Glass, NY)
As destructive methods, which use small samples taken from glass fragments ICP-MS, PIXE,
neutron activation analysis, scanning electron microscopy with energy dispersive X-ray emission
spectrometry (SEM/EDX), atomic absorption spectroscopy, Electron Microprobe Analysis
(EPMA), and laser ablation Fourier Transform Ion Cyclotron Resonance Mass Spectrometry
(LA/FTICRMS) can be used. Solid state nuclear magnetic resonance (NMR) can be a useful
technique to study the silicon local structure.
9
A few examples will be given in order to illustrate the use of some of the techniques in the
characterization of glass artefacts.
In recent work in our Laboratories, two glass production centres of the 18th century, Covina and
Marina Grande, have been studied. The provenance of glasses from excavations, as well as from
objects from national museum collections which seemed to be produced in Portugal, can only be
unequivocally determined through analytical characterization. The archaeometric studies of glasses
will have to deal first with the archaeological glasses taken from the sites where the glass
manufactures have been made. Excavations were already made in the site where the manufacture of
Coina was installed, and a large and important collection of glass samples were collected. The
samples were analysed in our laboratories using micro-EDXRF. A typical spectrum made using
micro-EDXRF is shown in Fig. 8.
Fig. 8. micro-EDXRF spectrum using a portable spectrometer ArtTAX – glass fragment of lead
glass of the “Real Fábrica de Vidros de Coina” indicating the elements Si, K, Ca, Mn, Fe, Sr and
Pb.
A preliminary analysis of 29 fragments showed that they could be divided into the following four
groups according to their chemical compositions: potash glass, soda-lime glass, high-lime low-
alkali glass and lead glass (Fig. 9).
10
Fig. 9. Plot of potassium and calcium oxides contents for the Coina fragments [13].
One of the fragments showing weathering, characteristic of archaeological samples which have
been in contact with water, was studied in detail. This fragment was from the archaeological Coina
excavation (Fig. 10) [14].
Fig. 10. Coina glass fragment: 1 - corroded surface; 2 - original green olive glass.
The use of SEM allowed the study of the different layers of corrosion as can be seen in the Figs. 11
and 12. The corrosion layer present fissures (or cracks) with two distinct phases, one heterogeneous
and the other homogeneous. After identification of the elements present using SEM/EDX, the
elements were chosen and a small area of the original fracture was scanned.
11
Fig. 11.SEM image of the glass fragment with increasing amplifications of a selected area; (a) area
observed (x10); (b) area analysed (x50); (c) different layers showing the glass corrosion (x200).
Fig. 12. SEM/EDX of the corroded glass fragment.
Just a few comments are made here. There is an enrichment of silicon and aluminium at the surface
as a consequence of the corrosion by water. Black conical inclusions observed at the surface have
12
an appreciable quantity of iron and manganese probably due to contamination from the soil. No
calcium was lixiviated which may be due to a surface rich in silica and alumina.
Ion beam techniques are currently used for the analysis of glass surfaces. It is reported here the use
of PIXE using a beam of protons of the Van de Graaff accelerator at the Portuguese Nuclear
Institute to study the yellow silver staining [15]. Fig. 13 shows the characteristic spectrum of the
yellow staining colour due to the presence of nanoparticles of silver.
200 400 600 800 1000
0
1
2
3
Fig. 13. UV-Vis spectrum of silver staining.
A typical example of the X-ray maps using PIXE obtained from a cross-section of a sample painted
with yellow silver staining is shown in Fig. 14.
Fig. 14. Elemental X-ray maps of yellow silver staining (106×106 µm 2 ) using PIXE.
It is possible to identify the simultaneous presence of Cu and Ag in the near surface region.
However, while the former is clearly confined to this region, the latter distributes also in the glass
bulk, indicating that diffusion of Ag into the glass matrix has occurred.
Another technique that has been explored in our Laboratories is Laser Ablation Fourier Transform
Ion Cyclotron Resonance Mass Spectrometry which can be used to characterize glasses [16]. One
interesting feature of this type of mass spectrometry is its high resolution, so isobars can be easily
distinguished. The positive ion spectra give the main constituents of glass exception made to the
13
boron as no positive ions can be observed. At the same time the observation of several clusters in
the negative ion spectra indicates the type of glass analysed. Figs. 15 and 16 show the mass spectra
of two different types of glasses, a soda-lime silicate glass and a borosilicate glass. As can be seen
the negative ion spectra are different for the two glasses. The negative ion spectrum of a soda-lime
silicate glass gives a series of clusters with silicon, oxygen, sodium and aluminium. The spectrum
of the borosilicate glass shows predominantly the ions BO2 - , BONaH
- and BO3H
be seen in the positive ion spectrum.
Fig. 15. Negative ion mass spectrum of a soda-lime silicate glass; the colours show the families of
clusters observed.
Fig. 16. Negative ion mass spectrum of a borosilicate glass
SiO2 -
(SiO2)O -
(SiO2)SiO2 -
(SiO2)2O -
(SiO2)2NaO-
(SiO2)3NaO-
(SiO2)4NaO-
(SiO2)5NaO-
Final Remarks
In this short review it was shown that there are several techniques which can be used for glass
characterization being their choice dependent essentially if destructive or non-destructive methods
are to be used. A few examples were given either of simple chemical analysis of glass objects or of
fragments having in mind provenance studies. In this short review no mention was made to the
isotope ratio analysis, in particular oxygen and lead, which can contribute with important
information in provenance studies. An example of the study of corrosion was given using
SEM/EDXF, a technique widely employed also for the study of archaeological glass.
Acknowledgments
We thank the “Fundação para a Ciência e a Tecnologia” for financial support and the Museum of
Glass of Marinha Grande for lending the ruby glass goblets. We are grateful to Stephen Koob from
The Corning Museum of Glass for providing us the photos for Fig. 7.
References
[1] James E. Shelby, Introduction to Glass Science and Technology, The Royal Society of
Chemistry, 1997.
[2] J. Pelouze, E. Fremy, Traité de Chimie Générale, Analitique, Industrielle et Agricole, Ed. Victor
Masson et Fils, Paris, 1865.
[3] M. H. F. Vaz Fernandes, Introdução à Ciência e Tecnologia do Vidro, Universidade Aberta,
1999.
[4] L. M. Liz-Marza´n, Nanometals: formation and colour, Mater. Today (February 2004) 26.
[5] Antonio Neri, Christopher Merret, The Art of Glass, The Society of Glass Technology, 2003.
[6] W. A. Weyl, Coloured Glasses, The Society of Glass Technology, Sheffield,
1951.
[7] F. E. Wagner, S. Haslbeck, L. Stievano, S. Calogero, Q. A. Pankhurst, K.-P. Martinek, Before
Striking Gold in Gold-Ruby Glass, Nature, 47 (2000) 691-692.
[8] C. Pantano, J. P. Hamilton, Rivista della Staz. Sper.del Vetro, 6 (2000) 81-86.
[9] C. Bray, Dictionary of Glass, 2nd Edition, A& C Black London and Univ. of Peensylvania
Press, Philadelphia, 2001.
[10] S. Koob, Conservation and Care of Glass Objects, Archetype Publications, London, 2006.
[11] G. Schulze, I. Horn, H. Bronk, “A new concept for the quasi non-destructive micro sampling of
historical glasses”, Fresenius J. Anal Chem, 358 (1997) 694–698.
15
[12] F. M. Lopes, A.. M. Lima, M. C. Freitas, Ho M. Dzung, A. Pires de Matos, II Jornadas
Nacionales sobre El Vidrio en la España Romana, Fundación Centro Nacional del Vidrio, Granja
de San Ildefonso - Segóvia, Spain, 8 -10 November 2007.
[13] F. M. Lopes, A. M. Lima, M. Vilarigues, J. Coroado, C. Carvalho, A. Pires de Matos, Real
Fábrica de Vidros de Coina – Chemical Analysis of Archaeological Glass Fragments, AIHV
Annales du 17 e Congrès, 2007.
[14] F. Lopes, Graduation Thesis, Universidade Nova de Lisboa, 2006.
[15] P. Fernandes, M. Vilarigues, L. C. Alves, R. C. da Silva, "Stained glass from Monastery of
Batalha: non-destructive characterization of glass and paintings", Journal of Cultural Heritage, 9
(2008) e5-e9.
[16] J. Marçalo, M. Santos, A. Pires de Matos, Characterization of Glasses by FTICR Mass
Spectrometry, Phys. Chem. Glasses, 43C (2002) 421-423.